This proposal presents a program of research that coordinates recent advancing in computational linguistics, artificial intelligence, and information science in the development of a basic research tool for investigating fundamental issues in natural language access to text. The research is motivated by the recognition that increasing amounts of medical information are available in text form and should be accessible through computer facilities. The specific objective of the research is the development of a prototype system that will allow physicians and other health professionals to access information in a computerized textbook of hepatitis through natural language dialogues. Natural language queries will be interpreted using a general natural language interface developed at SRI. A text access component will be developed that uses representations of the propositional content of text passages and of the hierarchial structure of the text as a whole to retrieve relevant information.